@shepai/cli 1.95.0 → 1.96.0-pr281.fcbc0cd

package/README.md

@@ -2,9 +2,9 @@
 
 # Shep AI
 
-### Multi-Session SDLC Control Center for AI Coding Agents
+### Autonomous AI Native SDLC Platform
 
-_Manage Claude Code, Cursor CLI, or any coding agent across the full development lifecycle — from requirements to deploy — with parallel sessions, instant context switching, and human-in-the-loop control_
+_Run `shep` in any repo to manage features through the full development lifecycle — requirements, planning, implementation, review, and deployment — using AI coding agents with human-in-the-loop approval gates._
 
 [![CI](https://github.com/shep-ai/cli/actions/workflows/ci.yml/badge.svg)](https://github.com/shep-ai/cli/actions/workflows/ci.yml)
 [![npm version](https://img.shields.io/npm/v/@shepai/cli.svg?color=cb3837&logo=npm)](https://www.npmjs.com/package/@shepai/cli)
@@ -17,367 +17,124 @@ _Manage Claude Code, Cursor CLI, or any coding agent across the full development
 
 <br />
 
-[Features](#features) · [Quick Start](#quick-start) · [CLI Reference](#cli) · [Architecture](#architecture) · [Contributing](#contributing)
+[Features](#features) · [Quick Start](#quick-start) · [CLI Reference](#cli-reference) · [Architecture](#architecture) · [Contributing](#contributing)
 
 <br />
 
-<img src="docs/poc/imgs/00-control-center-hero.png" alt="Shep AI Control Center — manage all features across repos from a single canvas" width="100%" />
-
-[![Live Demo](https://img.shields.io/badge/Live_Demo-shep--ai.github.io-blue?style=for-the-badge)](https://shep-ai.github.io/cli/poc/)
+<img src="docs/screenshots/features-guide/01-dashboard-canvas.png" alt="Shep AI — Control Center Dashboard" width="900" />
 
 </div>
 
-### Install
+---
+
+## Quick Start
 
 ```bash
+# 1. Install
 npm i -g @shepai/cli
-```
 
-### Try it
+# 2. Run Shep — completes setup on first run, then opens the web UI
+shep
 
-```bash
-cd your-project/
-shep feat new "Add unit tests for all source files"
+# 3. Browser opens at http://localhost:4050 — you're in
 ```
 
-<details>
-<summary><b>List features</b> &mdash; <code>shep feat ls</code></summary>
-<br />
+---
 
-```
-  Features for /home/user/projects/my-app
+## Features
 
-  ID        Name             Status          Branch               Created
-  1b0a7969  Add Unit Tests   ✓ Completed     feat/add-unit-tests  2/13/2026
-```
+- **Full SDLC automation** — AI agents handle requirements gathering, research, planning, implementation, and review
+- **Human-in-the-loop gates** — Approval checkpoints at PRD, Plan, and Merge phases (configurable or skip with `--allow-all`)
+- **Multi-agent support** — Plug in Claude Code, Cursor CLI, or Gemini CLI as your coding agent
+- **Web UI dashboard** — Interactive React Flow canvas showing features and repos as graph nodes with real-time SSE updates
+- **Git worktree isolation** — Each feature gets its own worktree, no branch switching or stashing
+- **CLI-first workflow** — Full feature management, agent control, settings, and tooling from the terminal
+- **TUI onboarding wizard** — Guided setup on first run
+- **Per-repo SQLite database** — All data lives locally in `~/.shep/`, no cloud dependency
 
-</details>
+---
 
-<details>
-<summary><b>Show feature detail</b> &mdash; <code>shep feat show 1b0a7969</code></summary>
-<br />
+## CLI Reference
+
+### Daemon
 
 ```
-  Feature: Add Unit Tests
-
-  ID           1b0a7969-b420-4754-832f-5c38801b4852
-  Slug         add-unit-tests
-  Description  Add comprehensive unit tests for all existing source files
-               in the project to ensure code correctness and improve test coverage.
-  Repository   /home/user/projects/my-app
-  Branch       feat/add-unit-tests
-  Status       ✓ Completed
-  Worktree     ~/.shep/repos/9bbd30de/wt/feat-add-unit-tests
-  Spec Dir     ~/.shep/repos/9bbd30de/wt/feat-add-unit-tests/specs/001-add-unit-tests
-  Agent Run    732a1fc2-a346-4052-a028-dcb1a46b6f2c
-
-  Timestamps
-  Created  2/13/2026, 12:15:08 AM
-  Updated  2/13/2026, 12:15:08 AM
+shep                                  Start daemon + onboarding (first run)
+shep start [--port <number>]          Start web UI daemon (background, default port 4050)
+shep stop                             Stop the running daemon
+shep restart                          Restart the daemon
+shep status                           Show daemon status and live metrics
+shep ui [--port] [--no-open]          Start the web UI in the foreground
 ```
 
-</details>
-
-<details>
-<summary><b>Isolated worktree per feature</b> &mdash; no stashing, no branch switching</summary>
-<br />
-
-Each feature gets its own git worktree — fully isolated from your main branch.
+### Feature Management
 
 ```
-~/.shep/repos/9bbd30de/wt/feat-add-unit-tests/
-├── .git                    # linked worktree (not a clone)
-├── src/
-├── tests/
-├── specs/
-│   └── 001-add-unit-tests/
-│       ├── feature.yaml    # status tracking
-│       ├── spec.yaml       # requirements & scope
-│       ├── research.yaml   # technical decisions
-│       ├── plan.yaml       # implementation strategy
-│       └── tasks.yaml      # task breakdown
-├── pyproject.toml
-└── README.md
+shep feat new <description>           Create a new feature
+      [--repo] [--push] [--pr]
+      [--allow-prd] [--allow-plan] [--allow-merge] [--allow-all]
+      [--parent] [--fast] [--model] [--attach]
+shep feat ls [--repo]                 List features
+shep feat show <id>                   Show feature details
+shep feat del <id>                    Delete a feature
+shep feat resume <id>                 Resume a paused feature
+shep feat review <id>                 Review a feature
+shep feat approve <id> [--comments]   Approve a feature
+shep feat reject <id> [--feedback]    Reject a feature
+shep feat logs <id>                   View feature logs
 ```
 
-</details>
-
-<details>
-<summary><b>Auto-generated spec files</b> &mdash; real output from a single <code>shep feat new</code></summary>
-<br />
-
-> **`spec.yaml`** — requirements & scope
-
-```yaml
-name: add-unit-tests
-oneLiner: Add comprehensive unit tests for all source modules
-summary: >
-  Add unit tests for every module in the package. The project currently has
-  minimal test coverage (only 7 tests total). This feature adds tests for
-  config, agents, flows, tools, observability, and CLI — using pytest with
-  mocking to avoid external dependencies.
-sizeEstimate: M
-openQuestions:
-  - question: 'Should we set a minimum coverage threshold and enforce it in CI?'
-    resolved: true
-    answer: >
-      Set 70% minimum via --cov-fail-under=70. Achievable and prevents regression.
-```
+### Agent Management
 
-> **`research.yaml`** — technical decisions with rationale
-
-```yaml
-decisions:
-  - title: 'Test Framework'
-    chosen: 'pytest (already configured)'
-    rejected:
-      - 'unittest — More verbose, no fixture system'
-      - 'nose2 — Unmaintained, no advantage'
-    rationale: >
-      Already a dev dependency with asyncio and coverage plugins configured.
-
-  - title: 'Mocking Library'
-    chosen: 'unittest.mock (stdlib)'
-    rejected:
-      - 'pytest-mock — Adds a dependency for marginal convenience'
-    rationale: >
-      patch() for module-level replacement, MagicMock for spec objects,
-      assert_called_with for verification. No new dependency needed.
 ```
-
-> **`tasks.yaml`** — TDD task breakdown
-
-```yaml
-tasks:
-  - id: task-1
-    title: 'Create conftest.py with shared fixtures'
-    dependencies: []
-    acceptanceCriteria:
-      - 'mock_settings returns a Settings instance with known values'
-      - 'mock_llm returns MagicMock(spec=LLM) for agent tests'
-      - 'tmp_repo creates a temp dir with git init and one commit'
-    tdd:
-      red: ['Write test_smoke.py that imports fixtures and asserts properties']
-      green: ['Implement conftest.py with all three fixtures']
-      refactor: ['Remove smoke test — subsumed by real test files']
-
-  - id: task-3
-    title: 'Test config/llm.py'
-    dependencies: [task-1]
-    acceptanceCriteria:
-      - 'test_create_llm_with_openai_provider: verifies model, temperature, api_key'
-      - 'test_create_llm_with_cursor_provider: verifies base_url included'
-      - 'test_get_fast_llm: calls create_llm with settings.fast_model'
-    tdd:
-      red: ['Patch LLM and get_settings, assert constructor kwargs']
-      green: ['Tests pass — they verify existing code with mocks']
-      refactor: ['Extract common patch decorators if repetitive']
+shep agent ls                         List agents
+shep agent show <id>                  Show agent details
+shep agent stop <id>                  Stop a running agent
+shep agent logs <id>                  View agent logs
+shep agent delete <id>                Delete an agent
+shep agent approve <id>               Approve an agent action
+shep agent reject <id>                Reject an agent action
 ```
 
-> **`feature.yaml`** — machine-readable progress (updated continuously)
-
-```yaml
-feature:
-  id: '001-add-unit-tests'
-  branch: 'feat/001-add-unit-tests'
-  lifecycle: 'research'
+### Repository & Session Management
 
-status:
-  phase: 'research'
-  progress: { completed: 0, total: 12, percentage: 0 }
-  lastUpdatedBy: 'feature-agent'
+```
+shep repo ls                          List repositories
+shep repo show <id>                   Show repository details
+shep session ls                       List sessions
+shep session show <id>                Show session details
 ```
 
-</details>
-
----
-
-Running AI agents is easy. **Managing many of them through a full SDLC is the hard part.** Shep gives you a control center that handles the entire lifecycle — requirements gathering, planning, implementation, QA, and deployment — across multiple agent sessions running in parallel. See what each agent is doing, jump into any session's web preview, code editor, or terminal, approve plans when agents need your input, and let them keep working while you move on. Plug in Claude Code, Cursor CLI, Gemini CLI, or any coding agent. Each feature gets its own isolated worktree and dev environment. You manage the lifecycle, agents do the work.
-
----
-
-## Features
-
-> 🚧 **Under active development** — Web UI screenshots are from the design spec. Core CLI (`feat new`, `feat ls`, `feat show`, `settings`, `ui`, `start`, `stop`, `status`) works today. The `mem` command is still WIP.
-
-<table>
-<tr>
-<td width="55%"><img src="docs/poc/imgs/01-features-control-center-overview.png" alt="Feature properties with attachments and metadata" /></td>
-<td width="45%">
-
-### Feature Properties & Attachments
-
-- **Rich feature detail panel** with name, description, and metadata
-- **File attachments** — requirements PDFs, mockup images, API specs linked directly to features
-- **Team & metadata** tracking for ownership and collaboration
-- **Dependency mapping** between features across repos
-- **Iterative refinement** — update properties as requirements evolve through the lifecycle
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/02-feature-card-action-toolbar.png" alt="Feature card with action toolbar" /></td>
-<td width="45%">
-
-### Feature Cards with Quick Actions
-
-- **One-click dev server** launch directly from any card
-- **Action toolbar** — GitHub, web preview, terminal, code editor at your fingertips
-- **Live progress** tracking with percentage and lifecycle stage badge
-- **Unique IDs** for instant cross-referencing across CLI and UI
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/03-requirements-discovery-validation.png" alt="Requirements discovery questionnaire" /></td>
-<td width="45%">
-
-### AI-Guided Requirements Discovery
-
-- **Smart questionnaires** that surface the right questions for your feature
-- **Categorized inputs** — business, technical, and UX problem classification
-- **Priority tiers** (P1–P3) with clear impact descriptions
-- **Domain tagging** to scope the affected areas of your codebase
-- **Human breakpoint** — agent pauses here until you finalize
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/04-technical-implementation-plan-review.png" alt="Technical plan review with architecture decisions" /></td>
-<td width="45%">
-
-### Technical Plan Review
-
-- **AI-generated implementation plans** with full architecture rationale
-- **Technical decision cards** — see the chosen approach and alternatives side-by-side
-- **Architecture pattern selection** (microservices, monolith, serverless) with trade-off analysis
-- **Approve & continue** or send back for revision — you control the direction
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/05-plan-review-tasks-and-artifacts.png" alt="Task breakdown and generated artifacts" /></td>
-<td width="45%">
-
-### Tasks & Generated Artifacts
-
-- **Auto-generated documents** — plan.md, tasks.md, data_model.md created from your approved spec
-- **Task breakdown** with dependency tracking and status badges
-- **Artifact browser** — inspect any generated document directly in the panel
-- **Draft indicator** — nothing ships until you explicitly approve
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/06-qa-check-review.png" alt="QA checklist with all checks passing" /></td>
-<td width="45%">
-
-### Quality Assurance Gate
-
-- **Automated QA checklist** — functional, performance, compatibility, accessibility
-- **Performance benchmarks** — API response time p95 thresholds
-- **WCAG 2.5** accessibility compliance verification
-- **One-click Pass QA** when all checks are green
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/07-dev-environment-web-preview.png" alt="Integrated web preview in dev environment" /></td>
-<td width="45%">
-
-### Integrated Dev Environment — Web Preview
-
-- **Live localhost preview** embedded in the control center
-- **Isolated ports** — each feature runs its own dev server
-- **Server status** indicator with start/stop controls
-- **Open in Browser** for full-screen testing
-- **PR context** — branch, worktree path, and PR link always visible
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/08-dev-environment-vscode-tab.png" alt="Embedded VS Code editor in dev environment" /></td>
-<td width="45%">
-
-### Integrated Dev Environment — Code Editor
-
-- **Embedded VS Code** with full editor capabilities
-- **Git worktree isolation** — each feature in its own directory, no branch conflicts
-- **File explorer** with project structure at a glance
-- **PR badge** with commits ahead/behind and changed file count
-- **Zero context-switch** — edit code without leaving the control center
-
-</td>
-</tr>
-
-<tr>
-<td width="55%"><img src="docs/poc/imgs/09-dev-environment-terminal-tab.png" alt="Terminal tab showing running dev server" /></td>
-<td width="45%">
-
-### Integrated Dev Environment — Terminal
-
-- **Full terminal access** inside the feature workspace
-- **Build output, test results, agent logs** in one place
-- **Pre-configured** in the feature's worktree directory
-- **Dev server output** — see Next.js/Vite/Express startup and compile times
-- **Three-tab workflow** — preview, code, and terminal without leaving the UI
-
-</td>
-</tr>
-</table>
-
----
-
-## Quick Start
-
-```bash
-# 1. Install
-npm i -g @shepai/cli
+### Settings
 
-# 2. Run Shep — completes setup on first run, then opens the web UI
-shep
-
-# 3. Browser opens at http://localhost:4050 — you're in
+```
+shep settings                         Launch setup wizard
+shep settings show                    Display current configuration
+shep settings init                    Initialize settings
+shep settings agent                   Configure AI coding agent
+shep settings ide                     Configure IDE
+shep settings workflow                Configure workflow
+shep settings model                   Configure model
 ```
 
-## Daemon Management
-
-`shep` runs the web UI as a background daemon so your terminal stays free. Use these commands to manage it:
+### Tools
 
-```bash
-shep start              # Start (or re-open) the web UI daemon
-shep stop               # Gracefully stop the daemon
-shep status             # Show live PID, port, uptime, CPU%, and memory
+```
+shep tools list                       List tools with install status
+shep install <tool> [--how]           Install a dev tool
+shep ide-open [--ide] [--dir]         Open IDE in directory
 ```
 
-## CLI
+### Other
 
 ```
-shep                                  Start daemon (onboarding on first run)
-shep start [--port <number>]          Start the web UI as a background daemon
-shep stop                             Stop the running web UI daemon
-shep status                           Show daemon status and live metrics
-shep feat new <description>           Create a new feature
-shep feat new --interactive           Interactive mode with human checkpoints
-shep feat new --allow-all             Full autonomous mode
-shep feat ls                          List features across repos
-shep feat show <id>                   Show feature details
-shep feat preview <id>                Preview feature dev environment
-shep ui                               Start the web UI in the foreground
-shep settings show                    Display current configuration
-shep settings agent                   Configure AI coding agent
-shep mem add <text> --global|--repo   Add agent memory
-shep mem ls                           List memories
-shep version                          Version info
+shep version                          Show version info
+shep upgrade                          Upgrade to latest version
+shep run <agent> [-p prompt] [-r repo] [-s]   Run an agent directly
 ```
 
+---
+
 ## Architecture
 
 Clean Architecture with four layers. Dependencies point inward — domain has zero external deps.
@@ -387,7 +144,7 @@ flowchart TB
     P["<b>Presentation</b><br/>CLI · Web UI · TUI"]
     A["<b>Application</b><br/>Use Cases · Orchestration · Ports"]
     D["<b>Domain</b><br/>Entities · Value Objects · Services"]
-    I["<b>Infrastructure</b><br/>SQLite · LangGraph · LanceDB"]
+    I["<b>Infrastructure</b><br/>SQLite · LangGraph · DI"]
 
     P --> A --> D
     I --> A
@@ -398,48 +155,76 @@ flowchart TB
     style I fill:#ede9fe,stroke:#8b5cf6,color:#4c1d95
 ```
 
+| Layer          | Path                                | Responsibility                                    |
+| -------------- | ----------------------------------- | ------------------------------------------------- |
+| Domain         | `packages/core/src/domain/`         | Business logic, TypeSpec-generated types          |
+| Application    | `packages/core/src/application/`    | Use cases, output port interfaces                 |
+| Infrastructure | `packages/core/src/infrastructure/` | SQLite repos, LangGraph agents, DI (tsyringe)     |
+| Presentation   | `src/presentation/`                 | CLI (Commander), TUI (Inquirer), Web UI (Next.js) |
+
 ### Feature Lifecycle
 
-Every feature progresses through a structured SDLC pipeline:
+Every feature progresses through a structured SDLC pipeline with 9 states:
 
 ```
-Requirements → Research → Review PRD → Planning → Review Plan → Implementation → QA → Approval → Live
-     ↑              ↑           ↑            ↑           ↑             ↑          ↑        ↑
-   Agent         Agent       Human        Agent       Human        Agent      Human    Human
-                           breakpoint                breakpoint               check    gate
+Started -> Analyze -> Requirements -> Research -> Planning -> Implementation -> Review -> Maintain
+                                                                                   |
+                                                                               (Blocked)
 ```
 
-Breakpoints are configurable. In `--allow-all` mode, the agent handles everything autonomously.
+Human approval gates are configurable at PRD, Plan, and Merge phases. In `--allow-all` mode the agent handles everything autonomously.
 
 ### Tech Stack
 
-| Component       | Technology                                                                              |
-| --------------- | --------------------------------------------------------------------------------------- |
-| Language        | TypeScript                                                                              |
-| Package Manager | pnpm                                                                                    |
-| CLI Framework   | Commander                                                                               |
-| TUI Framework   | [@inquirer/prompts](https://github.com/SBoudrias/Inquirer.js)                           |
-| Web UI          | Next.js 16 + React 19 + shadcn/ui ([@shepai/web](./src/presentation/web/))              |
-| Design System   | Storybook                                                                               |
-| Build Tool      | tsc + tsc-alias (prod), tsx (CLI dev), Next.js (web dev)                                |
-| Database        | SQLite                                                                                  |
-| Agent System    | [LangGraph](https://www.langchain.com/langgraph) (`@langchain/langgraph`)               |
-| Vector DB       | [LanceDB](https://lancedb.com/) (file-based)                                            |
-| Embeddings      | [Transformers.js](https://huggingface.co/docs/transformers.js) (local, multiple models) |
-| Testing         | Vitest (unit/integration) + Playwright (e2e)                                            |
-| Methodology     | TDD (Red-Green-Refactor)                                                                |
+| Component       | Technology                                                                |
+| --------------- | ------------------------------------------------------------------------- |
+| Language        | TypeScript (ES2022)                                                       |
+| Package Manager | pnpm                                                                      |
+| CLI Framework   | Commander                                                                 |
+| TUI Framework   | [@inquirer/prompts](https://github.com/SBoudrias/Inquirer.js)             |
+| Web UI          | Next.js 16 + React 19 + shadcn/ui + Tailwind CSS 4                        |
+| Graph Viz       | React Flow (XYFlow) 12                                                    |
+| Design System   | Storybook 8.x                                                             |
+| Build Tool      | tsc + tsc-alias (prod), tsx (CLI dev), Next.js (web dev)                  |
+| Database        | SQLite (better-sqlite3, per-repo)                                         |
+| Domain Models   | TypeSpec -> generated TypeScript                                          |
+| Agent System    | [LangGraph](https://www.langchain.com/langgraph) (`@langchain/langgraph`) |
+| DI Container    | tsyringe                                                                  |
+| Testing         | Vitest (unit/integration) + Playwright (e2e)                              |
+| Methodology     | TDD (Red-Green-Refactor)                                                  |
+
+### Supported Tools
+
+Shep can detect, install, and manage the following tools:
+
+| Category   | Tools                                                                             |
+| ---------- | --------------------------------------------------------------------------------- |
+| IDEs       | Alacritty, Antigravity, Cursor, iTerm2, Kitty, TMux, VS Code, Warp, Windsurf, Zed |
+| CLI Agents | Claude Code, Cursor CLI, Gemini CLI                                               |
+| Dev Tools  | Git, GitHub CLI                                                                   |
+
+### Web UI
+
+The web dashboard runs at `http://localhost:4050` and provides:
+
+- **Dashboard canvas** — Interactive React Flow graph with feature and repository nodes
+- **Feature drawer** — Tabs for overview, activity, approval, rejection, PR info, deployment, and timeline
+- **Create feature form** — Start new features from the UI
+- **Settings, Tools, Skills, and Version pages**
+- **Real-time updates** via Server-Sent Events (SSE)
 
 ### Data Model
 
 ```
-Repository ──┬── Feature ──┬── Plan ──┬── Task ──── ActionItem
-             │             │          └── Artifact
-             │             └── Requirement ── Research
-             └── Analysis docs (cached)
+Repository --+-- Feature --+-- Plan --+-- Task -- ActionItem
+             |             |          +-- Artifact
+             |             +-- Requirement -- Research
 ```
 
 All data lives locally in `~/.shep/`. Per-repo SQLite databases. No cloud dependency.
 
+---
+
 ## Documentation
 
 | Document                                           | Description                           |
@@ -453,7 +238,6 @@ All data lives locally in `~/.shep/`. Per-repo SQLite databases. No cloud depend
 | [Guides](./docs/guides/)                           | User guides and tutorials             |
 | [Development](./docs/development/)                 | Contributing and development setup    |
 | [API Reference](./docs/api/)                       | Interface and model documentation     |
-| [POC: Feature Flow](./docs/poc/)                   | Interactive feature lifecycle demo    |
 
 ## Contributing
 

package/apis/json-schema/Feature.yaml

@@ -89,5 +89,5 @@ required:
   - openPr
   - approvalGates
 allOf:
-  - $ref: BaseEntity.yaml
+  - $ref: SoftDeletableEntity.yaml
 description: Central entity tracking a piece of work through the SDLC lifecycle (Aggregate Root)

package/apis/json-schema/SdlcLifecycle.yaml

@@ -11,4 +11,5 @@ enum:
   - Review
   - Maintain
   - Blocked
+  - Deleting
 description: Software Development Lifecycle stages for feature progression